Typical contemporary digital electronic circuits are based primarily on transistor devices fabricated from semiconductor materials. In fact, the development of microfabricated bipolar and field effect transistors led directly to the modern development of digital electronic circuits and microprocessors [see texts by Horowitz and Hill, “The Art of Electronics”; D. V. Bugg, “Circuits, Amplifiers and Gates”′]. Digital electronics refers to circuits in which there are only two states possible at any point. Typically these states are set by gate circuits comprised of one or more interconnected microfabricated semiconductor transistors that can be in one of two stable states: saturation or nonconducting. Because the gates have characteristically high impedance, these semiconductor transistors (and circuits) use electrodynamic input and output in the form of digital voltage pulses. These semiconductor transistors and circuits are directly electrically connected together by some conductive material. The voltage states corresponding to saturation and nonconducting are HIGH and LOW, which represent, respectively, the TRUE and FALSE states of Boolean logic. These states also correspond to bits of information, typically HIGH represents a “1” and LOW a “0.”
The class of tasks in which the output or outputs are predetermined functions of the input or inputs is called “combinational” tasks. These tasks can be performed by semiconductor transistor gates which perform the operations of Boolean algebra applied to two-state systems. Combinational logic is basic to digital electronics. The three most popular semiconductor transistor logic families presently in use are Transistor-Transistor logic (TTL), Metal Oxide Semiconductor (MOS) logic and Complimentary MOS (CMOS) logic.
A disadvantage of such semiconductor transistors is the fact that their size and packing density is limited by the inherent physics of their operation, including thermal restrictions and density of charge carriers. Moreover, to implement the logic of even a simple single logic gate (such as an AND gate for example) in semiconductor digital electronics usually requires a circuit composed of several transistors [and possibly resistors and diodes] which takes up further space. Finally, to make semiconductor devices that are non-volatile—i.e., to retain a particular logical state—typically requires complex device logic support structures and/or operational characteristics.
The above considerations, and others well known in the art, restrict the packing density of semiconductor devices. Recently, a new magnetic spin transistor has been developed which can perform substantially all of the operations associated with semiconductor transistors. This new transistor, including its structure and operation, is described in detail in articles authored by me and appearing in IEEE Potentials 14, 26 (1995), IEEE Spectrum Magazine 31 no. 5, pp. 47-54 (May 1994), Science, 260 pp. 320-323 (April 1993), all of which are hereby incorporated by reference.
The structure, however, of this new magnetic spin transistor has prevented it use as a logic gate for performing digital combinational tasks. To date in fact, such magnetic spin transistors have been limited to such environments as memory elements, or magnetic field sensors.
Moreover, a major problem to date has been the fact that there has been no feasible or practical way to interconnect one or more spin transistors together. This is because, unlike semiconductor transistors, spin transistors are low impedance, current biased devices, which cannot be directly electrically interconnected from one to another by using electrodynamic coupling and the transmission of voltage pulses.
Finally, another problem with previously known magnetic spin transistors is the fact the output current of such devices has not be large enough to accomplish current gain. Lack of current gain is another reason why contemporary magnetic spin transistors cannot be successfully interconnected together to form digital processing circuits, because the output of one device must be capable of setting the state of another device, a condition that can be stated as requiring that device fanout must be greater than one.